Question on capacitors
Tenicus, Fri Jun 29 2007, 02:19PMIm building a MMC using 942C20P15K-F capacitors. The transformer is an NST rated at 10.5 kv, 30 ma.
I needed a total capacitance of around .0075 uF. Since each cap is .15 uF, twenty of them would get me that amount.
Is it OK to have only one string?
Also, do you divide the NST voltage by the capacitor's DC voltage or AC voltage when finding the correct # of caps?
Thanks
Re: Question on capacitors
GreySoul, Fri Jun 29 2007, 04:53PM
GreySoul, Fri Jun 29 2007, 04:53PM
Re: Question on capacitors
J. Aaron Holmes, Fri Jun 29 2007, 06:30PM
Right. Assuming you're using a static gap, I'd go with the "one string of 15" as for the 12/30 NST "Static Gap LTR" value. 7.5nF is a little too small. The difference between the LTR for a 10/23 and 12/30 is only one cap, according to this chart, so I'd pick the higher of the two. You're actually somewhere in the middle.
Cheers,
Aaron, N7OE
J. Aaron Holmes, Fri Jun 29 2007, 06:30PM
GreySoul wrote ...
Right. Assuming you're using a static gap, I'd go with the "one string of 15" as for the 12/30 NST "Static Gap LTR" value. 7.5nF is a little too small. The difference between the LTR for a 10/23 and 12/30 is only one cap, according to this chart, so I'd pick the higher of the two. You're actually somewhere in the middle.
Cheers,
Aaron, N7OE
Re: Question on capacitors
Sulaiman, Fri Jun 29 2007, 07:53PM
A string of 15 sounds ideal; (10nF 30kV)
You could consider a string of 15 that would allow you to test at 14 or 13 or even 10 in series,
each change will of course require re-tuning the primary, (and possibly the spark gap)
you may find that less than 15 is better
you've plenty of voltage margin to play with (up to full dc rating of capacitor at least)
If things get too warm or go 'POP' then you overdid it a bit
Increasing capacitance (once larger than resonant) actually reduces voltage stress on the NST
but I think you should maintain aim for one spark per 50/60 Hz half-cycle
For the topload you could consider this kind of approach
Sulaiman, Fri Jun 29 2007, 07:53PM
A string of 15 sounds ideal; (10nF 30kV)
You could consider a string of 15 that would allow you to test at 14 or 13 or even 10 in series,
each change will of course require re-tuning the primary, (and possibly the spark gap)
you may find that less than 15 is better
you've plenty of voltage margin to play with (up to full dc rating of capacitor at least)
If things get too warm or go 'POP' then you overdid it a bit
Increasing capacitance (once larger than resonant) actually reduces voltage stress on the NST
but I think you should maintain aim for one spark per 50/60 Hz half-cycle
For the topload you could consider this kind of approach
Re: Question on capacitors
Tenicus, Fri Jun 29 2007, 08:52PM
Beautiful.
Thank you 3 very much.
Tenicus, Fri Jun 29 2007, 08:52PM
Beautiful.
Thank you 3 very much.
Re: Question on capacitors
ArcLight, Fri Jun 29 2007, 09:26PM
My little 3.5†coil is powered by a 15Kv, 60ma NST. Originally, I used one string of ten CD-942 .15uF caps. However, I kept destroying caps (one at a time, not the whole bank at once) after only a few minutes of run time. So I decided instead to use two strings of 19 caps. No more blown caps even after several nights of hard running. So I believe that a single string will work fine at your power level.
Concerning your question about using the AC or DC rating on the cap. I have read every reference that I can find about MMCs on this website and all of the other websites that I could find. It seems that most people use the DC rating whether running DC or not when using the CD caps. However, as my experience above demonstrates even using a tight static gap and a safety gap something very bad was happening inside those caps.
I don't understand where you got the .0075uF number from, my calculations yielded a result of .012uF so one string of 12 or 13 would be where I would start.
Regards, ArcLight
ArcLight, Fri Jun 29 2007, 09:26PM
My little 3.5†coil is powered by a 15Kv, 60ma NST. Originally, I used one string of ten CD-942 .15uF caps. However, I kept destroying caps (one at a time, not the whole bank at once) after only a few minutes of run time. So I decided instead to use two strings of 19 caps. No more blown caps even after several nights of hard running. So I believe that a single string will work fine at your power level.
Concerning your question about using the AC or DC rating on the cap. I have read every reference that I can find about MMCs on this website and all of the other websites that I could find. It seems that most people use the DC rating whether running DC or not when using the CD caps. However, as my experience above demonstrates even using a tight static gap and a safety gap something very bad was happening inside those caps.
I don't understand where you got the .0075uF number from, my calculations yielded a result of .012uF so one string of 12 or 13 would be where I would start.
Regards, ArcLight
Re: Question on capacitors
Tenicus, Fri Jun 29 2007, 09:53PM
I used the equation found at
Tenicus, Fri Jun 29 2007, 09:53PM
I used the equation found at
Re: Question on capacitors
sparky, Sat Jun 30 2007, 02:50AM
I use 28 caps (at 1600VDC, 0.68uF each) by 4 sets in parallel for my capacitor bank.
These are awesome caps!! Never had one blow!
sparky, Sat Jun 30 2007, 02:50AM
I use 28 caps (at 1600VDC, 0.68uF each) by 4 sets in parallel for my capacitor bank.
These are awesome caps!! Never had one blow!
Print this page